Psychological safety has become one of the most celebrated concepts in organizational development. Unfortunately, it is also one of the most destructive. Not because the theory is inherently flawed, but because the implementation is so universally distorted that the theory no longer matters. What most organizations call “psychological safety” bears little to no resemblance to its original construct. If anything, it has become a mechanism for eliminating the very friction necessary for cognitive adaptation.
The damage is neurological.
The Neuroplastic Requirement
Human cognitive adaptation operates through prediction-error signals. When the brain encounters information that violates existing schemas (a failed assumption, a contradicted belief, a challenged perspective), it initiates a neurochemical cascade: acetylcholine signals attention to the mismatch, epinephrine heightens alertness, and dopamine reinforces corrective rewiring when adaptation succeeds (Huberman, 2021). This sequence drives synaptic strengthening, dendritic growth, and circuit reorganization, which are essential for learning and what neuroscientists term “functional compensation“: the brain’s reorganization in which undamaged regions assume the functions of impaired areas or inefficient pathways (Voss et al., 2017).
This process is not optional. Neuroplasticity thrives on error signals and meaningful stakes. The brain requires consequences to adapt, not the consequence of humiliation or punishment, but the consequence of being wrong, of encountering resistance, of having predictions violated by reality or by rigorous counterargument. It is cause-and-effect.
Chronic threat or excessive stress impairs this mechanism by sustaining elevated cortisol levels, which suppress hippocampal neurogenesis, reduce prefrontal flexibility, and promote maladaptive rigidity (McEwen et al., 2015). Moderate, controlled adversity, however, drives adaptation. The question is not whether challenge is necessary. The question is whether psychological safety, as implemented, preserves or eliminates it.
What Psychological Safety Was Supposed to Be
Amy Edmondson’s (1999) original framework defined psychological safety as “a shared belief that the team is safe for interpersonal risk taking” (p. 350). The idea was good and straightforward: reduce the interpersonal punishment for admitting errors, asking questions, or challenging assumptions. In theory, this would enable people to surface mistakes, voice concerns, and engage in the error-driven learning that organizational adaptation requires. In theory.
The theory assumed that safety would lower defensive behavior without eliminating accountability. It assumed that protecting people from social punishment for engaging in learning would not translate into protecting ideas from intellectual scrutiny. It assumed that organizations would distinguish between interpersonal threat (which should be reduced) and intellectual challenge (which should be preserved).
Well, unfortunately, that assumption was wrong.
What Psychological Safety Has Become
In practice, psychological safety has become a tool for ideological conformity. Rather than creating conditions in which interpersonal risks are safe, contemporary implementations create conditions in which certain intellectual risks are prohibited. Viewpoints are categorized as “unsafe.” Dissenting perspectives are labeled “harmful.” Challenging protected narratives becomes an act of aggression. The result is not safety. It is control and censorship.
Sure, rare pockets of authentic implementation may exist where leaders maintain Edmondson’s original distinction between interpersonal protection and intellectual challenge. However, these exceptions are statistically and practically irrelevant to the dominant pattern. When the vast majority of organizational applications corrupt a framework beyond recognition, the framework itself becomes the problem.
This produces a neurologically perverse outcome: high baseline anxiety about what you can say, combined with low cognitive challenge about what you believe. Individuals learn not to voice contrasting opinions, not to challenge consensus, and definitely not to engage in the behaviors that generate prediction-error signals necessary for cognitive adaptation. The mechanism mirrors chronic stress inversion. Initially intended to lower threat, the distorted version creates a different threat (social exclusion for nonconformity) while removing the adaptive friction that controlled stress provides.
Epistemic Rigidity as Institutional Policy
When environments protect certain conclusions from falsification, they prevent the schema violations that would trigger neural updating. The brain entrenches existing pathways because no error signal suggests they need revision. This is Epistemic Rigidity personified: the cognitive inflexibility that emerges when belief systems become immune to disconfirming evidence (Robertson, 2024).
Distorted psychological safety institutionalizes epistemic rigidity. By classifying certain ideas as too dangerous to examine and certain perspectives as inherently threatening, organizations eliminate the prediction errors necessary for cognitive adaptation. The neuroplastic cost is severe. Chronic low-error environments downregulate dopamine release associated with novelty and mismatch detection, while simultaneously elevating oxytocin-driven in-group bonding, which further entrenches tribal priors (Rozin & Royzman, 2001).
What results is not resilience but fragility: untested assumptions reinforced by selective exposure, intellectual atrophy disguised as protection. This has profound implications for individual and organizational outcomes. Organizations celebrate psychological safety while producing psychologically brittle individuals incapable of tolerating disagreement, processing criticism, or adapting to challenge.
The Empowerment Alternative
The solution is not to salvage psychological safety by finding its “proper” implementation. The solution is to abandon the framework entirely in favor of empowerment. This distinction may be subtle, but it is significant.
Empowerment means granting individuals the agency, responsibility, and authority to voice dissent, challenge consensus, and engage with ideas that contradict their own. It does not engineer comfort. It demands courage. The key is in the contrast. Instead of removing consequences, it introduces them. It does not protect people from intellectual friction. It equips them to navigate it.
The distinction is fundamental. Psychological safety, as practiced, asks: “How do we make it easier for people to speak?” Empowerment asks: “How do we develop people capable of speaking despite difficulty?” The first treats individuals as fragile. The second treats them as leaders capable of growth.
Neuroplastically, empowerment aligns with how brains actually develop. Controlled adversity, meaningful stakes, and error exposure drive synaptic reorganization. Empowered individuals encounter resistance, process feedback, and adapt. Safe individuals avoid resistance, suppress feedback, and stagnate. The former builds capacity. The latter atrophies it.
This aligns with the principle of Contrastive Inquiry, a deliberate practice of seeking and examining perspectives that contradict one’s current framework. Contrastive Inquiry generates the prediction-error signals necessary for neuroplastic updating without relying on environmental engineering to eliminate discomfort. The goal is calibration: exposing beliefs to resistance in conditions where learning, rather than self-protection, becomes the adaptive response.
Implications for Leadership Practice
For leadership practitioners, the implications are clear. Again, the key is in the contrast:
- Empower voices, do not engineer silence. Develop individuals capable of voicing dissent, not environments that remove the need for courage.
- Institutionalize devil’s advocacy. Make contrasting perspectives structurally required, not socially prohibited.
- Measure error exposure, not error avoidance. High-performing teams surface and address mistakes openly. Low-performing teams suppress them to preserve comfort.
- Audit for viewpoint diversity. If everyone agrees, someone isn’t thinking. If certain topics are unspeakable, the organization is not safe. It is controlled and censored.
- Reject the safety framework entirely. The theory may be sound, but the practice is so universally corrupted that defending the distinction is strategic naivety. Focus on empowerment, accountability, and adaptive capacity instead.
Remember that safety produces comfort, and comfort minimizes consequence. When the consequence is removed, the brain has little reason to adapt. Neuroplastic change depends on error signals, friction, and meaningful stakes. Environments engineered for psychological safety suppress those signals by design. Empowerment does the opposite. It introduces responsibility, exposure, and the necessity to adjust. Minds are refined by consequence, not by insulation. Development follows challenge, not comfort. Act accordingly.
Keep learning! Check out the article titled “Leadership Development Must Be Uncomfortable.”
References
Duhigg, C. (2016, February 25). What Google learned from its quest to build the perfect team. The New York Times Magazine. https://www.nytimes.com/2016/02/28/magazine/what-google-learned-from-its-quest-to-build-the-perfect-team.html
Edmondson, A. C. (1999). Psychological safety and learning behavior in work teams. Administrative Science Quarterly, 44(2), 350–383. https://doi.org/10.2307/2666999
Heterodox Academy. (2020). Understanding the campus expression climate: Fall 2020. https://content.heterodoxacademy.org/uploads/CES-Report-2020.pdf
Huberman, A. (2021). Huberman Lab Podcast. https://hubermanlab.com
McEwen, B. S., Nasca, C., & Gray, J. D. (2015). Stress effects on neuronal structure: Hippocampus, amygdala, and prefrontal cortex. Neuropsychopharmacology, 41(1), 3–23. https://doi.org/10.1038/npp.2015.171
Robertson, D. M. (2024). Reasoned Leadership instructor’s bible. DMR Publications.
Rozin, P., & Royzman, E. B. (2001). Negativity bias, negativity dominance, and contagion. Personality and Social Psychology Review, 5(4), 296–320. https://doi.org/10.1207/S15327957PSPR0504_2
Voss, P., Thomas, M. E., Cisneros-Franco, J. M., & de Villers-Sidani, É. (2017). Dynamic brains and the changing rules of neuroplasticity: Implications for learning and recovery. Frontiers in Psychology, 8, 1657. https://doi.org/10.3389/fpsyg.2017.01657